Key takeaways
- Views and likes are vanity metrics. The numbers that matter are view-through rate, conversion lift, cost per acquisition, and pipeline influenced.
- Decide the single business outcome a video should drive before you shoot, then instrument for it. A video with no defined job can't succeed or fail meaningfully.
- True video ROI = (revenue attributed to video − total video cost) ÷ total video cost, and cost includes distribution and internal time, not just the production invoice.
- Attribution is rarely clean. Combine UTMs, landing pages, view-through windows, and a 'how did you hear about us' field, and accept a defensible range over false precision.
- Video is a compounding asset, not a one-off campaign. Judge it over 6 to 12 months and repurpose the winners instead of chasing viral hits.
Why views and likes are the wrong way to judge video
A video with 200,000 views can be a total waste of money, and a video with 800 views can be the best marketing investment a business makes all year. Views measure how many people the algorithm showed something to. They say nothing about whether the right people watched, whether they cared, or whether a single one of them became a customer.
The trouble with vanity metrics — views, likes, follower count, impressions — is that they feel like progress because they go up. Leadership sees a big number, everyone nods, and nobody asks the uncomfortable question: did this sell anything? When the video budget comes up for review, "we got 200,000 views" is not an argument. "Video influenced $140,000 in pipeline last quarter" is.
None of this means reach is worthless. Awareness is real and it matters. The mistake is treating a reach metric as a success metric. Reach is an input. Revenue is the output. The rest of this guide is about connecting the two.
The metrics that actually map to revenue
A short list of metrics consistently correlates with whether a video is doing its job. Most of them sit downstream of attention and intent, not raw exposure.
- View-through rate (VTR) and average watch time — how much of the video people actually watch. If viewers drop off at three seconds, the content isn't landing, no matter how many impressions you bought. Watch time is the best single proxy for whether a video holds attention.
- Click-through rate (CTR) — of the people who watched, how many took the next step: visited the site, clicked the link, tapped the CTA. This is the bridge from attention to intent.
- Conversion rate — of the people who clicked, how many did the thing that matters: booked a call, filled a form, bought, subscribed. This is where video meets the funnel.
- Conversion lift — the gap in conversion rate between people who saw the video and people who didn't. This is the closest thing to proof that the video caused the outcome.
- Cost per acquisition (CPA) — total spend divided by customers acquired. A video with a low view count and a $40 CPA beats a viral one with a $400 CPA every time.
- Pipeline influenced — for B2B and higher-ticket sales, the dollar value of deals where video played a role in the buyer's journey.
You won't track all of these for every video. Which ones matter depends entirely on what the video was built to do, which is why goal-setting comes first.
Set a measurable goal before you shoot
The most expensive mistake in video isn't a bad edit. It's shooting something with no defined job. A video built to do everything does nothing measurably. Before a single frame is captured, you should be able to answer two questions: what is the one outcome this video exists to drive, and how will we know if it did?
Match the metric to the funnel stage. Different types of business video exist to do different jobs, and judging a top-of-funnel brand film by bottom-of-funnel conversion is how good work gets killed.
- Awareness videos (brand films, social hooks): judge on watch time, VTR, reach among the right audience, and lift in branded search or direct traffic.
- Consideration videos (explainers, case studies, product demos): judge on CTR, time on site after watching, and progression to a lead form.
- Conversion videos (testimonials, sales-page videos, pricing walk-throughs): judge on conversion rate, CPA, and revenue.
Write the goal down before production as one sentence with a target: "This testimonial should lift the pricing-page conversion rate from 3.2% to 4% or better." Now the video can succeed or fail on purpose, and you've told your production partner exactly what to optimize for — pacing, message, CTA placement — instead of "make it look good."
How to track video across the funnel
Tracking video comes down to instrumenting the path from watch to action. The mechanics differ by where the video lives, but the principle is the same: make the next step measurable.
On your own site
This is where you have the most control and the cleanest data. Set up video engagement events in your analytics — GA4 or similar — for play, 25/50/75/100% watched, and CTA clicks. Then watch what happens next: does the page with the video convert better than the version without it? A/B testing a landing page with and without video is the single most reliable way to prove conversion lift, and when the video is genuinely relevant to the offer, most businesses see a meaningful bump.
On social and hosted platforms
YouTube, Instagram, TikTok, and LinkedIn give you native watch-time, retention curves, and CTR data. Lean on the retention graph especially, because it shows you the exact second people leave. For traffic heading off the platform, tag every link with UTM parameters — source, medium, campaign — so the visit lands correctly in your analytics instead of getting lumped into "social" or, worse, "direct."
Attribution: connecting video to sales and leads
Attribution is where most people either give up or fool themselves. The honest truth: perfect attribution doesn't exist. Someone watches a testimonial on Instagram, forgets about it, googles you three weeks later, and books. No tracking pixel captures that cleanly. The goal isn't certainty. It's a defensible, directional picture built from several imperfect signals.
- UTMs and landing pages — dedicated tracking links, and where it's worth it, dedicated landing pages for a video campaign, so you can isolate the traffic and conversions it drives.
- View-through attribution — most ad platforms will credit conversions that happen within a window (say one to seven days) after someone watched but didn't click. Useful, but treat it as directional, not gospel.
- The 'how did you hear about us?' field — the most underrated tool in the stack. A simple open field or dropdown on your contact and booking forms catches the messy real-world journeys your pixels miss. When people write "saw your video," that's qualitative gold.
- Self-reported and sales-team notes — for higher-ticket sales, have the team log whether a prospect referenced content. It isn't tidy data, but it's real.
Combine these and you get a range, not a single number, and that's fine. "Video contributed to somewhere between 15 and 25 leads last quarter" is more honest and more useful than a falsely precise figure nobody trusts. Leadership respects a defensible range far more than a made-up decimal.
Benchmarks: what good performance looks like
Benchmarks are context, not scorecards. Your industry, offer, and audience move these numbers a lot. But if you have nothing to compare against, here's a rough sense of what "working" looks like in 2026. Treat them as directional starting points, not guarantees.
- Short-form social (Reels, TikTok, Shorts): three-second retention above roughly 50% and full-view rates in the 15 to 25% range on a well-hooked video are healthy. Watch the first three seconds obsessively — that's where most videos die.
- YouTube: average view duration of 40 to 50%+ on a two-to-five-minute video is strong, and a thumbnail CTR above 4% is solid.
- Landing-page and embedded video: a play rate above 50 to 60% of people who see it, plus measurable conversion lift over the no-video version, is the win to chase.
- Video ads: a VTR of 25 to 35%+ and a CPA at or below your other paid channels means video is earning its place in the budget.
The most important benchmark is your own past performance. Beating last quarter matters more than matching an industry average built on a completely different audience and offer.
How to calculate the true ROI of a video
Here's the formula. The part most people get wrong is the cost side.
Video ROI = (Revenue attributed to video − Total video cost) ÷ Total video cost × 100
"Total video cost" is not just the invoice from your production company. It includes distribution spend, the internal hours your team put into scripting, review, and posting, and any tooling. A $6,000 video with $4,000 of ad spend behind it is a $10,000-plus investment, and pretending otherwise inflates your ROI and burns your credibility the moment someone checks the math.
A worked example
Say you produce a customer testimonial for CAD $5,500 and put $3,000 of paid promotion behind it over a quarter — total cost $8,500. Through UTMs, a dedicated landing page, and your "how did you hear about us" field, you can reasonably attribute 12 new customers, and your average customer is worth $2,400 in first-year revenue. That's $28,800 in attributed revenue. ROI = ($28,800 − $8,500) ÷ $8,500 ≈ 239%.
One more honesty check: not every video needs to show direct revenue. Some exist to build trust or shorten sales cycles, and their return shows up as "deals closed faster" or "fewer objections on calls." Just decide that up front — see the goal-setting section — rather than reaching for a soft justification after the numbers disappoint.
Turning underperforming videos into learnings
Some videos won't hit their goal. That's not failure, it's data, as long as you actually read it. The retention curve tells you almost everything: a cliff at three seconds means the hook is wrong, a slow bleed means the middle drags, and a drop right at the CTA means the ask is unclear or unearned.
- Weak hook, decent body? Re-cut the opening and re-post. You rarely need to reshoot.
- Good watch time, low CTR? The content works but the call to action doesn't. Fix the ask, the placement, or the offer.
- Good CTR, low conversion? The video's promise and the landing page don't match. That's usually a page problem, not a video problem.
- Everything looks fine but no sales? You may be reaching the wrong audience. Check who's actually watching.
The businesses that win with video treat it as a portfolio. They expect some videos to underperform, they double down on the formats and messages that work, and they repurpose winners across platforms instead of chasing one-off hits. Over time the cost per result drops, because you're making fewer guesses. This is also why measurement should feed your broader plan — video shouldn't be a line item that stands alone, but part of how you allocate spend, which we cover in how much a small business should spend on marketing.
If you'd rather not build the measurement layer from scratch, that's a large part of what a good production partner should bring to the table. At Arctec AI we scope video production around the outcome first — the goal, the metric, the funnel stage — and tie it into the wider social and content strategy so you're measuring lift, not just collecting views. If you want a straight answer on whether your current video spend is working, get in touch and we'll walk through the numbers with you.